Abstract
The presence of metallic particles has been recognized as a dangerous threat in gas-insulated substation (GIS). Such particles are initially free and move toward higher electric field regions such as triple junction i.e., spacer-electrode-gas interface. However, once these particles reach the spacer surface, they adhere to the spacer easily due to electrostatic image forces. From insulation point of view, the triple junction is the weakest point in GIS. The presence of such metallic particles on the spacer surface deteriorates the insulation strength. Thus, in order to improve the reliability of GIS, it is important to identify the size and the position of the particle adhering to the insulating spacer surface. One of the most promising methods to carry out such identification is by recognizing the partial discharges (PDs) provoked by such particles. This paper is aimed to discuss the particle size and position estimation by using the PD patterns and statistical analysis. The PD patterns were acquired using IEC 60270 method. Measurements were made to determine various PD signals caused by particle of different sizes at different locations on the spacer surface. The acquired PD patterns were characterized by a number of statistical parameters. The results show that the implemented technique could be used to distinguish between various particle sizes and positions at different SF6 pressures with a fairly high accuracy.